Ways to organise references

I'm actively looking for better tools to use to organise my reading. I'm trying to solve three problems:

  1. Too often I remember reading something somewhere but can't find the reference to go back to that paper.
  2. Compiling the bibliography for a paper is laborious, and it seems to me that there ought to be a way of automatic all that categorising of information into 'author', 'title', 'date', etc.
  3. I'm really bad at keeping track of papers I haven't yet read but am interested to sooner or later.
Here's what I've tried so far:

Following Erin's example, I started a wiki. This is a nice concept for discussing ideas, and if I were to use it properly it would solve problems #1 and #3. The trouble is that I am lazy, and it doesn't automate #2, so it's not as extremely easy as I would like to add a reference.

Today I've started using CiteULike, which is an online service specifically designed for this purpose. It solves problem #2 admirably for a sizeable subset of papers, because if the article is from one of the sources in the list on this page, all you have to do is go from the article's abstract/details page to the post bookmark and it fetches the bibliographic data for you. For those sources it doesn't cover, there's a form to manually enter all the data, which is still marginally more convenient than writing it out longhand, and at least standardises the format.

For problem #3, CiteULike allows me to categorise a paper as either one that I've already read or with a relative priority. For problem #1, there is the ability to file papers under custom tags (just like, which may be familiar to more people), and to add a public or private comment to each citation, so I can remind myself of why I thought paper X was interesting. I wish it would automatically fetch the abstract, but it's little hassle to paste that in.

So this post is partly a plug for a service that I'm impressed with so far, with the caveat that I only started using it today, but I'm also interested in feedback: does anyone else use services for these goals that you think I should look at?

Oh, and my CiteULike library is browsable - I use the username I almost always use: eldan.


Trackback URL for this entry is:


Well, when it comes to the laborious task of writing bibliographies, I have in the past used EndNote. It's not free, and the dovetailing with MS Word would probably not be useful for you (you use LaTeX for papers, right? - the one you sent looked like that style).

I checked out your CiteULike page. I'm not sure whether I like it more or less than EndNote for organizational purposes. I do like that it's web-accessible, because it's annoying to only be able to find my references when I'm at my lab. The community aspect has a lot of promise, too... I may have to check out the "statistics and methodology in the social sciences" guys.

(Hee! Did I mention it is so cool to see someone in a CS department reading stuff by Esther Thelen? Awesome. It gives me hope for the future of science when interdisciplinary is more than a buzzword in the service of a grant...)

Posted: May 20, 2005 09:41 AM

Have you looked at RefWorks offered by the Kelvin Smith Library. It is similar to EndNote, but has the benefit of being totally web-based.

Look at the KSL announcement or the KSL RefWorks description page for more information.

Posted: May 20, 2005 01:38 PM

I've just recently started using LaTeX, mainly because I was unhappy with Word's management of numbering for all sorts of references (bibliographic, table numbers, section numbers etc.), but I'm still learning the system, so I'm far from wedded to it yet. Unlike most LaTeX users I know that Word can give me the quality of output and fine grained control that I want, so in the end I'm just going to use whatever makes organising the logical structure of a document easiest.

Of course, using a free and web-based solution has big advantages, and something I like about both CiteULike and RefWorks is that they support multiple output formats, so choosing one of them doesn't commit me to using a particular document editing system. This is particularly important as it's rather hard to co-write a paper if one author tries to use Word and the other LaTeX.

Brian: thanks for the information; RefWorks does look like a really nice service. My one worry is what will happen when I leave Case? Would I have a choice between paying a personal subscription to RefWorks or letting the database expire?

Of course, the ability to export the entire database in bibtex and/or EndNoteformat makes this much less of a problem, since pretty much any system I might want to use can import at least one of those two.

Posted: May 20, 2005 02:39 PM

Erin: I just completely ignored your last point. Sorry about that - it wasn't intentional. Anyway, I promise I'll finally get around to writing the "about me" post that soon, but basically I'm much more interested in the problems that psychology and biology look at than the ones CS is interested in. I just went into a CS program to learn more about the tools I wanted to use.

Posted: May 20, 2005 03:30 PM

RefWorks is available for individuals to purchase for $100/year. Of course as you mentioned, your records can be exported to another bibliographic management program.

One advantage with RefWorks (and EndNote) is that they are compatible with many of Case's research databases through Kelvin Smith Library.

Posted: May 20, 2005 03:33 PM

Someone just posted an entire essay about creationism as a comment to this blog. To the spammer: your domain is now blocked from posting here. Sorry if this seems a little heavy-handed, but that was not the right way to start a dialogue, and I'm determined to keep things on-topic and spam-free here. To everyone else: it was amusingly full of the easiest-to-debunk pieces of creationist BS, like the jumbo-jet-by-chance fallacy. I have a copy in my email inbox if you really want to see it.

Posted: May 25, 2005 10:31 AM

Post a comment